DriveTone: Car Sounds for EVs

DriveTone (for iPhone) makes driving EVs more fun with responsive, real-time engine sounds as you drive. Because sometimes you want to make a little noise.

Comment 1

Inspiration

As an EV owner, I love how quiet they are. As a sports car fan, I miss how the engine sound creates a real connection with the car and the driving experience. With all the hype right now about virtual engine sounds for EV’s, I had this crazy idea to try to bring that to any EV, without any hardware.

What it does

Based entirely on the accelerometer & GPS data from a phone, the app physically simulates a full working powertrain of a car, and connects that to a game-quality audio renderer for sound effects and haptics. The result is an amazing soundscape that dynamically responds to your driving.

How we built it

The app is a native iOS/Swift app. I am also presently learning to leverage AI to run "software factories", so the entire app was built with a custom software factory setup that I designed. The app also uses a custom diagnostic viewer and sound mastering studio.

I have a lot of experience working with phone sensors, sensor fusion algorithms, signal processing, etc. For the sound effects, I partnered with existing content creators that develop sound mods for car games like BeamNG, Forza, and Assetto Corsa to license their work for the built-in sound packs.

We then thoughtfully re-mastered the sounds for replay on a car stereo listening environment.

App telemetry is done through TelemetryDeck and we have an internal "feedback" system to collect feedback each time the app is used on a real drive.

Challenges we ran into

When the simulation gets out of sync with the car's real-life state, it breaks the illusion and is a bad experience for the users. It takes great care to get the right algorithms to nail the sensor fusion experience and tune the physical simulation.

Wireless CarPlay has great sound quality but unusable latency. The first version launched with a hard-block in place for wireless CarPlay. This has proven to be a huge hit for adoption.

We also learned that GPS signals are not strong in many cars (CarPlay gets around this issue by using the car's GPS antenna) and many users with Bluetooth had a terrible experience. This was also a huge hit to conversion.

As good as these sound packs sound in the game, when the sounds are used in everyday driving there is a very low margin for error. Small loop artifacts and interference patterns between layers stick out like a sore thumb. We had to develop a robust audio pre-processing & remastering infrastructure.

Also, every car has a different sound system, so we had to build a tunable haptics layer to make the haptics system work in a frequency that the user's car can actually deliver.

Accomplishments that we're proud of

Social Media promotion during TestFlight betas and post-launch has been a big success, with over 300,000 views across a handful of posts, yielding 600+ installs. This has been great for early feedback and prioritizing our roadmap. The engagement and excitement from users of the app really pushed me to work harder and faster to release updates.

What we learned

This is my first experience with B2C / entertainment apps. It is a very different dynamic than B2B SaaS! A lot of learning about influencers, consumer behavior/pricing approaches, and how to do well on social media.

What's next for DriveTone

Figuring out how to monetize the app -- thoughtful pricing, a solid GTM plan, and getting lots of user feedback! Then, a social-heavy rollout!

Built With

Share this project:

Updates

posted an update —

DriveTone got its first 3 sales today! 2 of them were 7-day trials of $20 annual that converted (all the others canceled...) and one was the $20 annual (no trial) we present after the user exhausts the free 30 minutes of full-featured test drive.

Feels very validating to know the higher price point has been used a few times!

Log in or sign up for Devpost to join the conversation.

posted an update —

We got our first trial conversion last night! Hard to tell exactly the conversion rate since it's so early and there is confounding TestFlight paywalls but it's around 1-3% to the trial for now, so that's a nice start.

Log in or sign up for Devpost to join the conversation.

posted an update —

I know from tests posts during our test flight period that single Reddit posts with a good hook and video of our app in use can generate 30 to 80,000 views, and so that is my primary launch focus, is using these videos as hooks in as many Reddit communities and social media shorts as possible.

I have been able to get 100 new installs since our PR launch just 18 hours ago.

Log in or sign up for Devpost to join the conversation.

posted an update —

The trial mechanism for my app is pretty tricky since the app can only be used in car, and people have very different driving habits.

My original plan was to offer people a 7-day free trial when they first download the app before they've had a chance to use it. And if they deny that, they get 30 minutes of test driving before the app will lock down to only 2-minute drives or offer to purchase an annual plan with no trial.

I've had 22 people reach that and no purchases, but 2 of the people reached out to me directly saying 30 minutes wasn't really enough to try it out the way they wanted because of their driving habits.

This early in the product launch, I'd rather have more people on top of funnel than fewer. And so I was able to use Revenue Cat to adjust my paywall that I use at that point in the life cycle, to swap the annual-no-trial for annual-with-7-day-trial. That was pretty cool to be able to do with RevenueCat without an app submission.

Hopefully those two users will now convert to the trial.

Log in or sign up for Devpost to join the conversation.

Submission history